Gasifier Experimenters Kit

ALL Power Labs: Tools for Power Hacking

How to Build a Gasifier

Technical Resources

overview - gallery - reactor options - manfacturing - news - GEKwiki - GEKforum - test reports - downloads - purchase

GEK Report #3: Temperature Profile Data for Air Preheating / Syngas Cooling

Data is finally in for GEK v0.9 internal reactor temperatures during normal run conditions. The graphics below give measured temps at all critical locations relevant to air preheating, combustion and reduction zones, and syngas cooling. The graphics index these temps to pictures of the GEK reactor, as well as CAD drawings of the same. I suggest you open both graphics to follow the comments.

Summary:

The GEK v0.9 air preheating / gas cooling architecture is raising 25C atmospheric air to @600C by the time the air reaches the nozzles at the combustion zone. In the opposite direction, the incoming air cools the output syngas to 100-175C before it exits the gas cowling.

Preaheating the air to 600C significantly reduces thermal load on the combustion zone, which would otherwise have to do the same air "preheating" before oxidation reactions can begin. 600C is above the auto-ignition temperature of most components of pyrolysis gas, so the air entering the combustion zone is already at reaction temperatures without heat input from existing combustion.

Eliminating the thermal drag of cool incoming air results in higher temps in the combustion zone, greater tolerance to moisture in the fuel, greater tolerance to high air humidity, and/or increased turndown ratio. It also increases combustion rate, which in the current test resulted in combustion finishing well before the reduction constriction-- suggesting we need to lower the nozzle height from the typical "textbook" dimensions.

Using the incoming air to internally cool the syngas greatly reduces the need for post-gasifier cooling. In our case, we've eliminated the external radiator entirely. The cyclone,packed bed filter and blower provide enough extra cooling to drop the final output syngas temp to 50-75C.

Full Report:

This data was collected and averaged over three separate run sessions of 2-3 hours each, in June and July 2008. The fuel was halved walnut shells at approximately 15% moisture content. The GEK was set up as an inverted V-hearth reduction bell, with 2.5" constriction. The heat exchange lines are 5/8" corregated SS lines of 4' length. More details on the heat exchange architecture is here: GEK Report #1.

The reactor was pulled at a flow rate that experience suggests is about its mid range of operation. This corresponds to 1-1.5" of H2O vacuum across the reactor (before the cyclone and filter). We cannot yet quantify this flow rate in volumetric terms, but in experience it has corresponded to running a gasoline genset at around 2kwe. (Apologies for annoying vagueness, we're still trying to acquire a full instrumentation suite)

We hard installed 6 type K thermocouples into the GEK to generate the temp data. We were careful to keep TCs in the actual gas streams and not touching pipe or vessel walls. We also tried to weigh the reactor during running so as to measure fuel consumption, but alas, success eluded us.

Here's some pictures of the process, along with the locals mulling over the situation (Bear Kaufmann, Charlie Sellers and Dennis).

Here's what we found . . .

Air In: Atmosphere to Nozzles:

The air preheating lines are raising the incoming air to 450-500C by the point where the air lines penetrate the bottom of the reactor and head upwards to the nozzles. This is adjacent to the base of the reduction bell, where the syngas exiting temp is 600-650C. This is around a 150C temp differential, which is likely about as good as we can hope for in a winding gas-to-gas heat exchanger.

The air temp raises another 100C or so between the reactor bottom and the insides of the nozzles. We measured 550-625C inside the nozzles. The highest peak was 650C. This temp increase is from heat contributed by the char and ash insulation around the reduction and combustion zones. I'm not clear the degree to which this heat input is an added drag on the combustion and reduction zones, or if it is mostly recovered heat already lost to the insulation. We'll have to measure the insulation temp around the air inlets to figure it out.

Remember, the point of this heat exchange architecture is to start mining heat from the syngas immediately after the reduction zone, only after all thermo-chemical work is completed. Mining heat from the combustion and reduction zone may sproduce a net zero sum heat budget in the end, but it could still reduce max temps achieved along the path, and thus impact tar conversion. Maintaining acceptable max temps is ultimately more critical than efficiency, given the extreme "inefficiencies" of tar in an engine . . .

Combustion Zone:

Combustion zone temps wandered in the 1000-1250C range. At other times i've seen it drop to 800C and rise to 1300C with extreme pull rates or fuel moisture variations. The moisture of the fuel actually seems to have more impact on combustion zone temp variations than the pull rate (within reasonable limits).

Reduction Zone:

The temp drop between the top and bottom of the reduction bell is not as great as I expected. The constriction at the top of the reduction bell runs around 700-800C. This suggests the combustion has already finished before the constriction and reduction has been underway for some time. This suggests the significantly preheated air is increasing the combustion rate so the "normal" distance from the nozzles to the reduction constriction for combustion to complete is now too long. Or in other words, we need to lower the nozzles to again place the finish of combustion right at the constriction of the reduction bell. (note the configuration help a simple TC provides).

The bottom of the reduction bell is consistently 600-650C. I was very surprised to see this temp stay steady across a very large range of pull rates. The only way I could get the bottom of the reduction bell above this temp was to overpull the reactor, which pulled the reduction down onto the grate and axially outward, typically burning off the gas cowling paint. I was baffled until I pulled out the Boudouard equilibrium reaction chart and discovered a "knee" in the graph right where my temp was stuck. 600-650C is "knee" in the boudouard reaction rate change, below which the reduction of co2 to co slows to at point which is not terribly relevant.

Clearly there is a wealth of configuration optimization and gas flow rate limits that can be derived simply from knowing the top and bottom temps of the reduction bell. More on this in the next report.

Grate to Gas Exit:

The syngas loses about 200-300C from the bottom of the reduction bell to the edge of the grate at the gas cowling. Temps here at the wall before the gas rises up the gas cowling wandered between 350-450C. I was surprised to see this significant of drop just from the reduction bell to the edge of the grate, as this is all of about 5" of travel. Another 200-250C or so is lost on the rise up to the gas outlet, resulting in 100-175C at the gas cowling outlet. (Post reactor cooling in the cyclone, filter and blower bring the final temp down to 50-75C, but none of this heat is recovered back into the system).

The air inlet temps show nearly the same changes but in the opposite direction. 25C at the air inlet. 300-350 at the bottom of the heat exchange tube spirals, where the grate intersects the gas cowling. 450-500C at the inner grate, next to the reduction bell where the air tubes turn up to penetrate the reactor bottom and go to the nozzles. And again, 550-625C at the nozzles.

Through the entire air in and syngas out path, the differential temp seems to hover around 150C. This does not suggest the heat in the syngas is equal to the thermal sink of the incoming air. I seem to remember the syngas out has about 40% more heat capacity than the incoming air, due mostly to volume differences (can someone clarify the specifics here?). This suggests the syngas is a better air preheater than the incoming air is a syngas cooler. The only reason the syngas out gets as cool as it does is from the added heat lost to the surrounding gas cowling and other surfaces.

This excess of sensible heat in the syngas out vs the air in also suggests the heat exchange system will be relatively tolerant of less-than-optimal designs, and still result in near max possible air temps at the nozzles. To the degree the heat exchange efficiency is improved, there is additional capacity we can use to do other heating work, while still ending at the same temps at the nozzles, and also further cooling the output gas.

What's Next:

I'm interested in using this excess heat exchange capacity to support the addition of water/steam, and/or IC exhaust to the incoming air. To the degree we can add h20 or co2 to the incoming air (while maintaining acceptable temps to consume tars in the combustion zone) our gas quality will go up, while our biomass fuel consumption will go down. Gas will come out slightly cooler too.

Thus starting with the v1.0 GEK, I increased the SS lines from 4' to 6'-- even though the 4' length as tested seemed to already produce the max incoming air temps one could expect across a small gas-to-gas heat exchanger. The v2.0 run will similarly have 6' heat exchange lines, along with an improved spiral configuration at the bottom of the reactor.