[Stoves] Purpose of testing and value given to char left.

Mon Feb 24 13:13:15 CST 2014

Dear Frank

You have captured this correctly:

>I think we are all in agreement with Crispin the final purpose is to save
trees and to Ron it's important to place a value on the char left over. 

They are two different questions and they have two different answers. Both
have value.

>You suggest we determine the wood used for each burn and after several
burns we have the average wet wood used. 

Actually we have a protocol for this and it is included in the Test Method
<http://tungkuindonesia.org/images/downloads/CSI_Indonesia_Test_Methods_2014
-1-14_v7.pdf> . 

>It will take many Runs (I suggest  ~ten statistically?) 

We are using three.

>And then we go to the next house and how many more houses, all using the
same stove (the variable), to come up with a statistical value.  

This is not actually necessary. There are ways of using focus group to make
a determination that is accurate enough to make a comparative assessment.
Remember something very important that changes the number of tests required
(a lot):

We are seeking comparative fuel consumption, not absolute numbers. This can
be shortened to a comparative test, not a certified result.

At the moment we are using a comparative test to pretend the result is
'certified', meaning 'absolute' but in reality, the lab and the procedure
and the number of replications, even the calibration protocols and lab
training would not really allow one to claim with precision that the result
was 'certified' correct. We have to do what we can with the equipment
available, in the time we have and in a spirit of fairness. 

It is easy when we are not trying to get a 5% or 10% improvement. As with
the stove replacement programme in Mongolia, we only allow into the
programme stoves that are at least 90% better than the baseline product.
That is a huge difference so there is no quibbling as to whether the product
is 15% or 12% better. It has to be 90% to 98% better to be interesting to
us. 

In the final analysis, the exact reduction a stove gives, say 21% v.s. 25%
improvement, is not as important as being sure that lots of people adopt a
better stove, and they are happy with it and use it most of the time. Being
really sure it is 24% better and only used 30% of the time is not really
what we want to see in the final analysis.

For this reason the social science surveys that were conducted were really
good, detailed examination of what is cooked, and when, by whom and so on. 

>Then we give all these same houses a different stove for testing. Is this
your plan?  

Nope. That is not practical. The bottom line is the products have to be much
better on emissions, reasonably better on fuel consumption, and bought
because people want to own and use them. In other words high adoption.
Turning that into numbers is always a nightmare and we do the best we can.
In this case it is with a new protocol that is strongly informed by local
patterns of use. That should give us more predictable results in the field.
Because the social science (social impact potential) has been assessed
beforehand, we can make a pretty good estimate of whether or not a stove
will be acceptable. One metric is cooking power, another is controllability.
A third is lighting speed. Another is cost and durability.

>A lab test using fuel Energy as a proxy reduces the variables such that a
lot of tests can be done in a short time at a reasonable cost. Like was done
before. The biggest variable is the left over energy after the WBT ended and
this can be accurately done using a calorimeter on a dried and ground sample
-as done at Aprovecho OR my past suggested E450c method. 

One of the very confusing things that has permeated the stove testing
community is the erroneous idea that the heat value in the fuel left over
has a great deal of significance. It has significance if you want to know
the heat transfer efficiency. It does not have a lot of significance if the
fuel is not useful in the next fire. For example, you might accurately
determine the heat value of tiny chips of charcoal in the ash. And what
would you do with het number? It would not tell you anything about how much
fuel it uses because that 'fuel' is a 'mechanical loss' (meaning energy lost
without burning it).

Fuel consumption is not determined by measuring the energy content of fuel
left over. It is measured by determine the amount of fuel needed each time
the burn cycle is completed. That is all. It is really a lot simpler than
people have made out. 

>As for the value to be placed on the char left over: The quality of this
char ranges from toxic to plants -to- sucking up nitrogen -to- adding heavy
metals -to- producing a large plant and doubling production. To place a
value one needs to collect a quantity and determine its value -if any for a
purpose. 

Well, it might have all sorts of values, but it is not relevant to the
amount of fuel needed to start and run the stove the next time. Because the
question we have been asked is 'How much fuel does this stove use, on
average, in that community?" we should answer it to the best of our ability.

>.For marketing we may be able to state that stove such-and-such
consistently puts out a char of quality to improve soils, sequester carbon
or can be used for char cookers -after collecting a composite of the char
from a stove and testing it.  

That is all interesting information, supplying answers to different
questions, not the one that is most frequently asked.

Regards

Crispin

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20140224/3d3b4d83/attachment.html>