[Stoves] A wisdom of Rebecca's stove

[Crispin Pemberton-Pigott]

Dear Ron

 

>The following is an attempt to be sure that char-making stoves get a fair
shake in your proposed (but still unknown) new method of stove testing.  

 

I am pretty sure you will defend the interests of the char making stoves. To
do that you are going to have to stay tuned to the purpose of testing an
individual stove. What happens outside that stove is not part of that
stove's emissions or efficiency rating. Such external considerations are
happily included in programmatic analysis. I have said this before and there
is not going to be another version of it: What a stove does with the fuel
fed into it has to be quantified separately from the other things that come
before and after. That applies to fuel preparation and fuel remnants. All
that external information is real and can be considered by the person making
a rational choice about how a stove fits into a broader system.

 

>>'Wasted' is not a term of criticism, it is used to mean 'not used' and is
a 'loss'. Losses classically are divided into 'chemical losses' (in the form
of un-oxidized gases) and 'mechanical losses', meaning unburned fuel that
falls to the ground from a conveyor, so partially burned fuel in the ash and
char remaining is in the latter category of mechanical losses.

   RWL1:   When reporting on stoves intentionally designed primarily for
making char, how about replacing "wasted" with "char produced"?   Since the
char may be used again for many purposes, what is wrong with always
reporting char production and let later users of the test report decide how
to use the data?

 

That would perhaps mislead many people and create room for baffle-gab and
misdirection. Even if there is char deliberately produced, not all the char
is useable and it is in any case a sub-section of the remnants/outcome of
combustion. As noted before there is moist fuel, dry fuel, torrefied fuel,
partially burned fuel, charred fuel and ash. There is always some carbon in
the ash. It is difficult to find out exactly how much is in each category;
it is usually not necessary to report more than the fuel consumption or the
system efficiency which is what the great majority of questions are about.
We are adding the heat flux rate because it is so valuable as a screening
method.

 

The ash also has many purposes. If you wanted, we could report the mass that
looks like char and the mass that looks like ash. That, as you say, lets
people decide afterwards what the implications of them are. I think the more
important point as Jim made clear, is that the energy value of those
remnants is not important if the remnant is not fuel for the stove being
rated. I believe that was specifically directed to this and several similar
questions.


>>The question that arises, as highlighted by Jim, is whether or not the
fuel can be burned in the same stove during the next replication. If it can,
then it is unburned fuel that will go to the next round and could therefore
be deducted from 'fuel consumed'. If not, then it falls into the category of
'mechanical losses' and has been 'consumed'. That means for fuel consumption
purposes it should not be deducted from the mass of fuel consumed per
replication.

    >[RWL2:  Same issue.  This seems to imply that char will either have to
be consumed or only be considered a loss.  No third option?   Was this your
intent?

 

It is not just the char. Anything left over that cannot be used in the stove
on the next replication is a loss, whether we dress up the language or not.
Some stoves cannot even use totally dried wood.  The intent is to report the
performance of the stove being tested. Nothing else.  Combustors and heat
exchangers have been tested for centuries and it is strange to see the slow
speed at which very ordinary methods are applied to domestic stoves by those
most intimately creating and evaluating them. Doesn't that strike you as
odd? 

 

>    [RWL3.   For others,  Jim introduced a new term of "apples" for
non-char-making stoves (and I thought was correctly done - really no
difference from the old method) and "oranges" for char-making stoves (the
only subject of the webinar).   

 

It is of course a little more complicated than that. All stoves burning
biomass make some 'char'. The problem was that the char was being
arbitrarily assigned a heat value that was probably not correct much of the
time and also the stove was being credited for char produced as if it was
unburned raw fuel when reporting the fuel consumption, which clearly it is
not. That needed to be corrected. 

 

The fact that some stoves are making huge amounts of char (relative to the
total energy available to start with) brought this to a head when stoves
were claiming a 'fuel efficiency' of 50% while consuming as much or even
more fuel than the baseline stoves which were 15% efficient. Clearly
something was wrong with the calculation method. There was. We are agreed on
how to correct it. Jim covered it very well.  It would be difficult to claim
that a stove that consumes 50 kg of fuel a week does not, when it does.

 

As to what happens to the char produced, that is outside the stove, but is
part of the total energy system. It is quite reasonable to examine what
happens to the leftovers. It is also correct to examine the energy it takes
to prepare or transport the fuel in the first place. Fuel is part of a
larger system.

 

>I take this to mean that Crispin's "new method" is not going to be
implemented by Jim.

 

He was in fact explaining the new method to you - that is not new.  I was
referring to a new method of generating three representative replications. I
have already discussed it with Jim. I have also discussed it with others. It
is a great simplification for the tester and provides the unskilled tester
or those with only a scale and a thermometer an opportunity to make a pretty
accurate determination of a stove's performance. The replication idea is
from the SeTAR Centre. There is a definition of 'fuel consumption' that goes
with it, clarifying what has been murky - by which I mean the claims for
fuel consumption that were based on heat transfer efficiency and energy
efficiency, rather than the need for more fuel. I have settled on 'system
efficiency' as perhaps the best way to describe the raw fuel in and the work
performed out. That ratio is the system efficiency. 

 

 I am still not understanding either what Crispin's new method is or what is
wrong with what Jim proposed (which also is "new"), and hope Crispin will
explain what is wrong with either/both.

 

I am extending the changes past the agreed reinterpretation of the fuel
consumption number. What Jim and I agree already (he wrote about it some
time ago you may recall and it was discussed here as well) is that we have
to examine what happens to fuel remaining in order to assess the fuel
consumption. If the stove can't burn it in its normal operation how can it
be considered fuel for that stove? If it is fuel for a different stove, test
that on too and made a programmatic decision.

 

There is of course the complication of stoves that can generate char in one
mode and burn it in another. That has not been resolved yet. It could be
that for three replications we should have to perform several iterations
that used the fuel in various ways to create a realistic performance
assessment. It still boils down to how much new fuel the stove needs from
the available supply, on average, per replication.

 

>>For stoves that cannot burn any of the fuel remaining, the current methods
over-report the system efficiency and under-report the fuel consumption. It
is because of this that we are implementing a completely new approach to the
determination of fuel consumption. 

   The heat transfer efficiency question is also being 'spruced up'.

    [RWL4:   In the first sentence, assuming that Crispin's  "cannot"
includes "is not designed to",

 

If it is designed not to, then it cannot, right? 'Cannot' is the right word.
We are concerned with the question of whether or not the stove can use the
mass remaining - whatever it is. In a stick fed stove like an open fire one
can argue that the sticks (now dried and blackened) can be used in the
subsequent fire. No problem. That is what people do to save fuel. At the end
there will be some more of the same.  Perhaps we will find that some TLUD's
can burn anything that is not pure char at a loading of 10% or 40%.  I was
intrigued by the idea that we might mix char from pellets into new pellets
and burn some of the char. Some stoves might be able to use less fuel that
way. We have to consider all design and operational possibilities.

 

Regarding the heat transfer efficiency, to get the closest number you can,
the pot material and mass must be considered, the unburned CO too, and the
lid should be on the pot. It should not have a floating insulator on the
water because that insulates the pot and falsifies its actual performance.
That is what I meant by sprucing up the method. The purpose of the HTEff
number is to assist designers of heat exchangers, so it should be a close to
the actual number as possible. I have another completely different approach
I can explain at another time. Important: the heat transfer efficiency does
not represent the fuel consumption. Relative heat transfer efficiencies do
not represent relative fuel consumption.  Keep that in mind. 

 

>I do not understand why the present formulae used by Jim/EPA/GACC in any
sense of this phrase, does "over-report the system efficiency and
under-report the fuel consumption"    

 

This refers to stoves that produce char and cannot burn it.  If you pretend
the char is unburned fuel, you understand the fuel mass burned.  The two
phrases amount to the same thing. If you under report the system efficiency
by saying is used more fuel than it did, the follow-on is that the fuel
consumption is reported to be higher than it actually is. This inverse is
also true. If you deduct the char remaining from the fuel consumed, it
pretends that the char remaining is unburned fuel that could be used on the
next replication. If that is true, no problem. If it can't burn it, then it
is not true. That was the entire point of the apples and oranges analogy.

 

>I believe Jim, GACC, and the whole stove community would greatly appreciate
hearing what is going to be done by Crispin in the new Indonesian tests.  

 

I have given descriptions of a couple of the elements of the assessment
method. One is the heat flux rate - the rate in Watts that a stove delivers
heat into a pot's heated surface. This will be assessed. If it does not meet
the minimum requirement, it means people are unlikely to want to buy it
because it will be viewed as underpowered. We don't want to waste time
assessing stoves we are pretty sure no one in the target community will
want. You might make a wonderfully efficient stove that has no emissions but
which is so low in its maximum cooking power that it will be rejected. We
can tell what the user wants by determining the heat flux rate for stoves
they say are acceptable (not even 'good', which would be a higher number).

 

Will a char-making stove have its char output reported in any sense?  

 

It will be reported as a mass of 'what looks like char'.  The reason is that
there will be a small bonus (as previously reported) for stoves that make
char while saving fuel through a higher system efficiency.

 

Suppose we have a minimum system efficiency of 45% for a particular
application, and suppose you have a device that is 65% efficient (about the
same as an LPG stove).  That is quite a bit of fuel saved. If you were to
turn that saving into a lower system efficiency (45%) plus produce some
char, the stove will be credited with a slightly high figure than the 45%.
It is a bit complicated to explain here (sliding scale) but it boils down to
a 10% credit if you can meet the system efficiency requirement and make char
at 25% (as defined in the text).  I believe this is the first time such a
credit has been offered. The purpose is to encourage the use (probably as
fuel, but at least for some economic benefit) of the char produced. This is
a programmatic decision, incorporated into the rating system. It does not
make the stove more fuel efficient and it does not state that the stove is
more fuel efficient than it actually is. It is just a way of rating the
'pass/fail' question about fuel efficiency and char making.

 

>What is the new formula to correct this major claim on inaccuracy?  What is
the "spruce up"?   Can all of Jim's present set of numbers be inferred from
the proposed new test method output?]

 

There are problems remaining with the WBT and my opinion on them is in the
document analysing it in detail (2008). Some issues remain uncorrected. The
use of the mass of water remaining when boiling is one. It causes needless
variability that should not be there.  The use of the mass of water
remaining when simmering is another and is probably unresolvable. The mass
of water in the pot has virtually no effect on the fuel consumed or the
emissions during simmering so it does not matter what the mass is. Thus no
'specific' performance metrics should be calculated for simmering (now
re-named 'low power' though it is not low power either). 

 

The main numbers to report for any stove are the fuel consumption and
emissions per useful MJ in the pot. Simmering does not have useful MJ in the
pot because ideally, there is no change in enthalpy so no work is done.
Ergo, no useful MJ.  The consideration of the mass of the pot and its
material needs to be incorporated. The British, Indian, SeTAR, Chinese and
Indonesian tests already consider it. The Indian version of how to do so is
the best and the SeTAR method is copied from it. It is in turn descended
from an old British standard. The Chinese method was developed independently
I think.

 

Because of the interest that is emerging in the char making ability of some
stoves, we will provide a space to report it but will not make its
characterisation part of the stove performance report which is confined to
fuel consumption and emissions per MJ in the pot(s). I am sure others will
take an interest in it later.

 

I think all of Jim's numbers could be inferred but he is generating numbers
based on the old WBT 3.1 data sheet some of which are agreed not to be
valid. For some reason they are still there. One is the efficiency of
simmering which you can still see on the new spreadsheets copied from the
old ones. It is misleading and untrue, as the WBT method does not determine
the heat transfer efficiency when simmering. It just reports a number based
on the missing water which may or may not indicate something. Having the lid
off makes it worse because of a pretty large IR radiation error (on the
order of 200 Watts).  As mentioned before, the specific emissions and fuel
consumption when simmering are also not informative so we are not generating
them. The heat transfer efficiency is compensated for chemical losses as per
standard boiler analysis. The fuel consumption figure is compensated for
mechanical losses (as Jim agrees is necessary). 

 

So we are pretty much on the same page, or getting there bit by bit. There
are a lot of resolution issues (precision and accuracy) which the new
triplicate replication method goes some way to improving. It happens to be
easier and simpler, but if it was not we would still do it because it gives
a more precise result. I say that because making additional replications of
a WBT4 - Jim mentioned some large numbers in the webinar - does not improve
the precision of the result.  It only increases the confidence of where the
middle of the error bars are. The size of the error bars is inherent in the
protocol and the instruments.  We have to develop methods with smaller error
bars. Biomass stoves are notoriously difficult to characterise well.
Probably the most difficult. Anything we do for them will automatically
improve the results for other fuel-burners.

 

Regards

Crispin

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20130902/61b7c754/attachment.html>