[Gasification] RPS experience with linear hearth
Peter & Kerry
realpowersystems at gmail.com
Wed Feb 1 21:58:20 CST 2012
Hi Doug & list,
I have moved the discussion to a more appropriate message header that
better reflects the content. Also the messages so far skip across a lot
of ground and it is too easy to get confused trying to cover multiple
threads of discussion within one reply, more so when there is no
differentiation in the text style between individual contributors.
In this message I will only focus on the gas out of the RPS linear
hearth, it is already long winded but am attempting to provide as much
detail as possible without compromising proprietary bits.
Doug has mentioned our gas analysis as being "very unusual" for an air
drawn system so for the benefit of the list I provide it here so people
can see for themselves:
Major Gasses,
Hydrogen: approx 36.00%
Carbon Monoxide : approx 28.00%
Nitrogen: 29.00%
Carbon Dioxide: 6.81%
Low level Gasses,
Oxygen: Less than 0.02%
Methane: 0.96%
Argon: 0.37%
There were also a range of minor gases in the less than 50ppm range.
The flow rate measured at the time of the test was 130m3/hr, though
this was not recorded on the lab certificate.
This is not full flow and the same system has been measured up to
400m3/hr without apparent over aspiration for a 20 minute run before
overheating of the fan motor caused it to trip. The sustained upper
limit has not been determined and may well be lower or higher (it has
taken a while to get a suitable high temperature fan of adequate
capacity, but will have one within the next few weeks, like most
components we have ended up building this ourselves), but the system is
quite comfortable at 200m3/hr with similar gas quality observed in the
flare and can be turned back to 40m3/hr without losing this.
Now at this point of our development we are not making any loud noises
about this, other than simply reporting it in a couple of local industry
presentations and now here. Up until recently it was only a single
analysis result. Instead we have largely kept our heads down and tried
within our own limited financial resources understand what is happening,
identify the feed stock parameters involved and get it further validated
when we could.
We have made numerous attempts to engage university researchers to
formally measure system performance without any real success. We had
enormous difficulties in early days trying to get gas analysis done,
with some amusing moments, like delivering a 4 litre gas sample bag of
raw unfiltered gas taken hot from our development unit to the chemistry
department of a local university because they indicated they could
analyse it, then getting an excited phone call rattling off a long list
of compounds they found, none of which included hydrogen or carbon
monoxide gas... It turned out that they had released the gas and opened
the bag to scrape a tiny amount of condensate from the inner
sidewalls...and ran the analysis on this. The (uncertified) condensate
analysis provided to us is nonetheless interesting and is copied below:
Condensate test results RPS first development unit
Percentage Compound
40.625 Pyridine C5H5N
2.298 Column Bleed
7.509 Phenol C6H6O1
0.750 Methyl Phenol C7H5O1
1.305 Methyl Phenol C7H5O1
3.340 Naphthalene C10H3
0.529 Dodlecene C12H24
1.163 1 Methyl Naphthalene C11H10
1.097 2 Methyl Naphthelene C11H18
0.566 Tetradecine C14H28
5.092 Biphenylene C12H8
0.562 2, 3 Dimethyl 1 Naphthelene C12H12
0.869 Dibenzofuran C12H8O1
0.674 ?
1.125 Fluorene C13H10
0.818 ?
0.552 ?
9.291 Anthracene C14H10
2.077 Anthracene C14H10
0.777 Anthracene C14H10
1.626 4HCyclopentaphenathracene C15H10
0.806 1methylAnthracene C15H12
0.951 2Pheny1Naphthalene C16H12
6.725 Fluroanthene C16H10
1.762 Pyrene C16H10
7.112 Fluroanthene C16H10
100.001
Doug also indicated that if we are only getting low overall condensate
levels then the water is probably going out with the gas as steam, and
ordinarily I would agree, except none of our observations of our linear
system whilst operating on optimal fuels support this.
The condensate analysis above does not show any free water at all (we
did query this at the time and asked whether this result was after water
had been excluded, ie reporting only the percentages of the non water
component, but were told no, if water had been present it would have
been reported).
Before I go on I would add that yes we have seen wet gas out of the
system, but only when running truly excessively high mc feed stocks in
the range of 30-50%. At 40% mc H2 drops to as little as 5% and CO to 11%
with a corresponding increase in CO2 & N2 (lab analysis result during
testing of mixed wood chip/ sewerage sludge blends). Stretching a length
of paper towel over the (un-ignited!) gas stream under these conditions
results in it getting rather damp quite quickly, and a brown condensate
dripping off the outer rim of the flare head can be observed (no funny
comments about the possible relationship to sewerage sludge please...).
Under other much less extreme gasification conditions though no moisture
collecting in the paper or free liquids on nearby metal surfaces are
readily apparent.
The following additional observations are for "chunky" wood fuels below
25% mc (the fuel spec at the time of the formal gas analysis, piece
sizes ranging from 25mm to 50mm on a side).
* Yes we do have gas cooling (of our own design like the rest of it),
and gas exit temperature immediately prior to the flare head are between
40oC and 70oC, depending on flow rate, and the current system also
includes mesh mist filters on the exit from the coolers, which we
thought might also be an efficient way of trapping the sub 10 micron
particulates, assuming that these would be wetted by condensate. You
would therefore expect to see a higher level of condensate than we are
getting, and I would have thought at least some steam visible in the
plume on contact with cold air (we have run it at 0oC air temperatures
in the middle of Winter without seeing such a plume).
*The formal gas analysis is consistent with a gas where a high
proportion of the available moisture in the feed stock is being cracked.
* The system "chuffs" when the mc is below 25%, a resonance coming from
the intakes sounding a little like a fast revving steam engine and the
upper hopper vibrating like a long, low drum roll.... Hand held digital
anemometer also records this as a regular, fast pulsing of the air
intake flows.
This seems consistent with a rapid cycling of water cracking and the
free oxygen made available displacing that from the incoming air. This
cracking uses up thermal energy which then drops below the threshold
required to support this water shift reaction, and the process pauses,
reverting to pulling in outside air to satisfy the oxygen demand,
temperature rises again and the process repeats. This has an immediate
knock on affect as for each air derived oxygen molecule displaced then 4
nitrogen molecules don't make it into the system, lowering dilution from
this source and raising overall gas quality.
* The reported gas quality is roughly mid way between a conventional
down draft and a indirect pyrolysis system, which tends to support the
above hypothesis as internal conditions cycle between air supported
combustion and thermal mass as the gasification heat source. The big
difference being the methane content, again consistent with more oxygen
being available in the air drawn system, the preferred reaction then
being Carbon + Oxygen, rather than Carbon + 4H2.
* We have seen a similar result when adding high oxygen content fuels
such as glycerine (a triple alcohol group) to the wood chip feed, the
flare quality and volume was maintained even though intake air flow
rates dropped to <25% of their normal range!
*Simple mass balances taking into account gas flow rate (intakes and
exit pipe), reported gas quality and measured fuel inputs closely agree.
*Charcoal from the ash bin has a very high fixed carbon in the 85-93%
range (reported by a NATA certified lab) consistent with high
temperature. Soot taken from the particulate collection system below the
cyclones has been examined with a microscope and we are told it had a
crystalline structure normally also only found when forming under high
(>1000oC) temperatures.
My wife Kerry, equal co developer, has asked that I also point out when
we designed the original system we made allowance for running it in
either downdraft, or updraft, mode and have used this facility to break
up bridging on occasion when working with difficult fuels (though on the
early units this does interrupt the gas flow to the flare). The current
(Mark 3.2?) under construction allows for this without stalling the
system. This was done as part of improving the overall material handling
side on the path to a more easily automated commercial model.
Cheers,
Peter
More information about the Gasification
mailing list