[Stoves] Ulaanbaatar Air Quality
Crispin Pemberton-Pigott
crispinpigott at gmail.com
Fri Nov 8 06:52:27 CST 2013
Dear Alex
Very good question and complex to answer so I will have to put down some
rules.
Lets define ignition as the first 30 minutes and the pyrolysed state
(poorly defined) as the second important point in the measurement.
The 30 minute window is the time taken to get the thing lit well and running
at an even power level.
Lets take the traditional stove as the baseline. These pics are quite high
resolution so you can zoom them.
The purple line is the burn rate so when it stabilises at a constant rate
(straight, angled line) the power level is constant.
The yellow line shows the accumulated PM to that point. So after 30 minutes
it has not quite stabilised but is at 35 minutes. It does not last long
before the burn rate starts to drop. This is typical of the product.
The initial smoke for 7 minutes is the wood lighting with maybe or maybe not
a couple of pieces of coal being affected. The huge blast of PM is the coal
being put on top of the wood fire (roasting it) or else it is being lit
crossdraft meaning the fire is hitting a pile of coal next to it. This is
the main portion of the total PM.
After it gets going, there is very little smoke. The second puff is the
refuelling of a hot stove at 70 minutes. It has a much shorter stabilisation
time and a similar burn rate profile.
I would say that the ignition is responsible for more than 90% of total
emissions, if you include refuelling.
Now lets look at a clean igniting stove (this is not a TLUD).
First note that the vertical scale is not quite the same.
This is a crossdraft stove with a hopper. The wood ignition can be seen
starting at minute 4 then dropping to nearly zero then the coal starts and
is burning very well by minute 25. The burn rate is constant by 30 minutes
indicating that the fire is established and constant. At minute 78 when the
traditional stove needed refuelling, it was only necessary to shake the
grate which made the spike in PM (ash, maybe a little BC).
This particular stove went through 6 more iterations until the smoke was
pretty much undetectable after 12 minutes. I would say that the ignition
period was responsible for more than 95% of total emissions.
Finally a TLUD (I have to hide the product name in this case) :
First note that the reported emissions rate is per net MJ, not per MJ as is
the case in the older tests above.
The ignition goes well with the wood (almost always larch or in some cases
pine). There is some smoke as the coal lights on top, largely due to the
cold combustion chamber and an excess of air in the chamber.
The increase in smoke at 55 minutes is not caused by someone fiddling or
refuelling. It is a problem found with many TLUDs that are created without
adequate testing. There is a lot of heat generated in the combustion chamber
and the primary air supply is not controlled well enough to limit the power
adequately. It heats the stove body and chamber (cast iron typically) so
much that at some point the pyrolysation rate exceeds the available air
supply. This took 55 minutes with this stove. As the gas production rate
starts to overwhelm the air supply combustion quality deteriorates and it
starts to make more smoke than it did during ignition. This is a common
problem and nearly no one is listening to us so we are starting a stove
design training centre to bring real time measurements to the attention of
the producers.
I would say that in the case of TLUDs the major part of the emissions comes
when the pyrolysation reasons the bottom of the pile and it flashes the
remaining volatiles into gas over a 15 minute period. The effect is common
in all TLUDs that burn to the bottom. There is no new raw fuel below to
cool the hot zone with new moisture. So it becomes in effect a self-heating
retort with thermal runaway.
Only a few TLUDs have proper and controllable secondary air. Some have it
but it is not preheated. The TLUDs with a cold supply of air at the top
blowing across the top of the chamber have the worst results of those with
secondary air. There is still a chance that the effect above will emerge.
The problem at root is a runaway pyrolysation effect caused by trying to
have the fuel be part of the air control system. If the fuel is not exactly
right there will inevitably be negative consequences. In other words the
stoves are too simple to work properly except in certain circumstances.
The call to chip fuel to a particular size and density is in part to keep
the stove simple. It is OK if you can organise it.
So what does a really clean lighting stove look like?
This is again a cross draft stove with a hopper:
This is the result of 18 months of work on ignition and fuel feeding
processes. It has a small amount of wood in the combustion chamber that is
lit on top to get the initial smoke out of the way and then it lights to the
side and runs constantly thereafter. It can be refuelled at will without
consequences and in fact this stove was refuelled. You cant see where. The
change in power was caused by shaking the grate.
It is with some annoyance that I see these stoves were specifically excluded
from the Berkeley comparative assessment of stoves they wrote for the GACC
on the excuse that people should not burn coal (which is what is says in the
introduction). There is a whole group of stoves which would all sit in the
bottom left corner of the stove performance chart deep into the Tier 4
section cleaner than anything burning wood. In fact some of these stoves
can burn wood very well too.
The final example (remembering to correct for the different metrics) is 1293
times cleaner than the baseline for a +99.9% reduction in PM2.5. That is
something to be praised, not hidden. What is the problem with letting people
know that coal can be burned this cleanly? A modern power station cannot
deliver electric heating and match the PM or CO numbers per net MJ. Those
producing stoves of this calibre are in Turkey, Mongolia and China. There is
a Japanese stove I have not seen which may also be in this league.
Incidentally, all the best stoves (the four approved for subsidy this year,
for example) produce no smoke after a certain time of running and starting
removing ambient PM, meaning they have negative emissions for at least part
of the time. The best ones, perhaps half, are able to scrub the air entering
the stove completely of all particles even if the ambient air has 200 or
more µg/m3. They literally clean the air to zero PM while burning wet
lignite in a natural draft 10 kW fire.
So far three types of combustors have achieved this: downdraft (BLDD),
end-lit down-angled crossdraft (ELCD) and updraft (TLUD).
Regards
Crispin watching the early morning snowfall
Crispin,
This may be an affront to your technical cred, but if you could humour me
with an educated guess, I have a question. How much of the
improvement/reduction in emissions is due to 'top ignition', better start up
and how much is due to improved steady state combustion?
Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20131108/844d2cfe/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.jpg
Type: image/jpeg
Size: 27149 bytes
Desc: not available
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20131108/844d2cfe/attachment.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image005.jpg
Type: image/jpeg
Size: 23100 bytes
Desc: not available
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20131108/844d2cfe/attachment-0001.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image009.jpg
Type: image/jpeg
Size: 26626 bytes
Desc: not available
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20131108/844d2cfe/attachment-0002.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image012.jpg
Type: image/jpeg
Size: 31011 bytes
Desc: not available
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20131108/844d2cfe/attachment-0003.jpg>
More information about the Stoves
mailing list