[Stoves] FW: ETHOS 2013: Where is the New Data on Stove Performance in the Field?

[Crispin Pemberton-Pigott]

Dear Paul

 

I really appreciate the care you have taken over the past month to
understand some of the intricacies of the stove testing issues that need our
collective attention.  I know you have been concentrating on combustion and
hardware issues more than test calculations.

 

>1.  The FULL data from the Quad 2 tests were provided to stimulate
discussions, and I am delighted that Crispin and some others are digging
into it as an example of the testing procedures.   

 

Without access to the raw data and the spreadsheet it is really not possible
to get much. There are different versions of the analysis sheets and one has
to first look to see what version of which test spreadsheet is in front of
you before you start looking for meaning in the numbers. Just last month I
was presented with a test result from a country to the north which was done
recently using WBT 3.0. That was a surprise. We do not really know much as a
result. Plus the CO emissions was given in ppm with no compensation for
dilution so we have no idea what the CO mass was, nor the combustion
efficiency. We have a long way to go.

 

>I do not run the rests and do not participate in the definition of the test
procedures.   

 

This is what makes the example so helpful. A char making TLUD enhances the
obviousness of the most of the errors which were always there, but which are
not as large for 'whole fuel burning' stoves. I did not conduct the test and
I had no say in which version of the spreadsheet was used. The test was paid
for, and the test protocol was followed well. Did Paul get his money's
worth? If not, is the fault of the lab? I think not. 

 

>Maybe other stove manufacturers could release (or point to current
availability of) full results of tests on their stoves.  

 

That would be very helpful. I have had difficulty obtaining the spreadsheets
used. Only a snapshot of final numbers are provided. I should point out for
the record that Paul provided the original spreadsheet immediately and a
copy of the test report. The spreadsheet is the version uploaded to the
pciaonline site after the beginning of June and before the version dated the
19thof June.  As far as I can tell, it does not contain any errors I know
of.

 

For those wishing to check which (recent) version they have, look for the
following:

 

May 2012 (or so) the water mass in the pots 2, 3 and 4 in the Hot Start
section have the formulas in lines AE12 and AL12 down to AE14 and AL14
reversed. This is only noticeable if you were using more than one pot.

 

Early June 2012, that error was corrected.

 

19 June 2012, the two tabs "TEST 1" and "TEST 2" are reversed. There is an
incorrect '2' in the formula in RESULTS i30 that should be a '3'. If you
download the file fresh, now, you should correct the formula in case someone
uses the outputs of that line.


>2.  I accept Crispin's position about the difference between FUEL
efficiency and ENERGY efficiency, but with this observation:      Crispin
said something like:  "Fuel from the woodpile" or "wood from the fuel pile"
should be what is counted.   TLUDs take in fuel and operate best if the full
batch of fuel is pyrolyzed, leaving behind char that is certainly not like
the original fuel.   That original fuel is gone!!

Your considered opinion on this point is well taken. When drafting a new
protocol it is this sort of first principles that should apply. Jim also
agrees on this point, if you have read his contribution.


>Therefore, if "fuel put into the stove, minus that SAME fuel that can be
saved" is the amount considered in the FUEL efficiency, then a TLUD could
use more fuel that other stoves.   

 

I would not at this time like to generalise. It is quite possible that a
TLUD making significant amounts of char could do a very respectable job
compared with an open fire. To paraphrase, "I have not yet begun to design!"
J  Let us not assume that future stoves will necessarily perform the way
current ones do.

 

>3.  This could imply that the stove was operated until the end of pyrolysis
even if that was many minutes after the complete WBT was conducted.   

 

I am not making that assumption tough it is one possibility. For example a
stove may have a fire shut-off mechanism capable of extinguishing the flame
within a short time. I do not want to anticipate what might be invented.
Because the nature of a test to get a particular metric requires the fire to
be stopped at the end, it is reasonable to make a plan to do so. There is a
method called the Burn-Out Test which was specifically designed to test
stoves with pellet-like fuel that cannot easily be shut down and which in
any case contain all sorts of half-burned bits of fuel. Dr Taylor rated that
method at about 15% error.

 

PARTIAL SOLUTION that creates havoc is that the SAME stove could be operated
identically with the amount of fuel carefully calculated to have pyroylsis
end within a minute after the WBT was completed.   

 

Although this is possible, it is not necessary. The performance of any task
like stir-frying, water heating or fish drying can be measured in a couple
of ways. I won't bore you with them now.  The penalty for an
un-extinguishable or uncontrollable fire is real. They are, after all,
dislikeable traits.

 

>Fill a TLUD with 3 kg of fuel for one test run, and then do it with 1 kg
for the second test run.   

 

This needs to be considered. I don't see the obvious advantage through this
explanation.

 

>In other words, Crispin's correct statement (that the  original loading of
1600 grams of wood) is NOT related to the amount of FUEL used up until the
time of the completion of the WBT.

Actually, it is. 

 

The test of the Quad 2 was conducted in 2 phases wherein the raw fuel was
loaded twice (the amounts are indicated) and both burned well past the
completion of the relevant section. The test as not completed in one go, in
other words. The remaining fuel was discounted to obtain the fuel burned
heat value (for getting the thermal efficiency, fire to water). The total
raw fuel needed to conduct a complete WBT I calculated from the mass of raw
fuel burned for each section and the char produced in that process. The
average for the three tests was 1550 g of raw fuel with a moisture content
of 15%.  If the stove was loaded with 1600 g it would in in all likelihood
complete the hot start and simmer sections in one go.

 

>4.  Because TLUDs can utilize so many different types of dry biomass
EQUALLY WELL, and because reduction of burning of WOOD is so important, then
what if the test was done with maize cobs or with stalks or with briquettes
from true "waste" or with inedible Jatropha seeds?    Then the following
would be true:
        "No fuel was taken from the woodpile"    and    "No wood was taken
from the fuel pile."

>So the impact of WOODY fuel savings means the TLUDs are the clear
winners!!!  

Nice try, but no! The fuel pile includes all biomass, not just wood. It is
quite reasonable to point out what fuels can be burned to potential
customers. They may have a lot of non-woody biomass available.


>Remember, the definition that the fuel is to be some uniform type of WOOD
(to ensure comparability between the stoves) is what has forced the TLUDs to
be tested with wood.  

 

There is no such definition. Anything that is biomass is included in the
definition of a biomass fuel. We are not trying to cause problems with
certain stove types. It is just that we are speaking from a first principles
perspective. Anything that can burn is fuel. Just about any biofuel can be
burned in a variety of stoves if mixed with something else. When the mass of
fuel burned is recorded, the test does not care what the source is, only the
operator. A test must be 'fuel blind' save for the mass, the moisture
content, the LHV and the ultimate analysis. It is people who make political
decisions about fuel types, not the test.

 

>How well I remember the Stove Camps at Aprovecho (I have attended seven
thus far) where the kiln-dried Douglas fir was the only acceptable fuel for
the testing.

That was an excellent example of why the emissions of a stove must be
obtained using the fuel that is appropriate and typical of the
implementation arena. Further, the burn cycle should be typical of the
intended use. Cecil Cook will be providing much more on this topic between
January and March 2013.


>5.  About testing of char-making stoves, that include the TLUDs:   
>Granted that char is not the original fuel.
>But the remaining char can have substantially greater value on a weight
basis (and perhaps in other ways also) than the biomass that was the
feedstock.   And that char can be for A) burning or for B) biochar (soil
amendment) or for C) carbon sequestration/climate change/carbon credits.
AND B and C can be cumulative!!!!     Or D)  it could be thrown away.

These are good 'social' or 'political' points that are not related to the
engineering metrics of a stove. The test methods must remain aloof from
ideologies. The reason is that if people think the 'fuel consumption' number
represents the fuel consumption of the stove, then we owe it to the people
to ensure the label matches the facts.


>So, concerning FUEL effeciency, W H O    R E A L L Y    C A R E S    what
the numbers are from the WBT xxx.yy?

Policy managers care because they are directing the distribution of very
large sums of money and they have a fiduciary responsibility to ensure they
are getting what they pay for. That is why I spend so much time trying to
correct the obvious and subtle errors on the test methods. I am not asking
everyone to get involved. It can be pretty boring. But anyone making
performance claims should understand the implications and limits of those
claims.


>.but "do not crucify the TLUDs on a cross of carbon."

 

That is a good line, but still a diversion. When you have a TLUD that makes
char and still outperforms a three stone fire, a built in stove and a
charcoal brazier, you will be singing a different tune. All that is required
is to keep your eye on the ball and not misrepresent the performance (if
possible).


>Fix (adjust, not "fix" as in illegal horseracing) the tests any way that
allows for true differences between stoves to be observed.    

 

I have been using the term 'hardened'. The 'true differences' have to be
rooted in defined metrics and standard units if it is to be broadly
accepted. Having 5 or 50 or 500 people agree to use a certain test method
does not make it a valid one. A test method might have dozens of correct
procedures and calculations but include one fatal flaw that invalidates the
final result. The dozens that are correct do not confer salvation.

 

>Just please do not have testing that poisons the waterhole for those who
are working with TLUD and other micro-gasifier stoves.

Well, that is well put. The current WBT calculations do exactly that. There
are several 'fatal flaws' which means the final number(s) misrepresent the
true performance of the product. Some stoves are accidentally 'optimised' to
perform well on a given test but this does not improve their actual
performance when viewed through the lens of a more rigorous test.


>Professional chemical engineer Hugh McLaughlin and I are examining the
issues of testing, and output (if any) should be ready before ETHOS in late
January.    

 

That is really good news. Please don't mix the political and social side of
fuel source and external matters into the test procedure. A stove works how
it works and consumes what it consumes. A well designed, precise and
accurate test brings the unvarnished truth. Don't shoot the messenger.

 

Regards

Crispin in the city of gold (goud as in Gauteng)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20121208/973f131a/attachment.html>