[Stoves] Heterogeneous testing protocols

Crispin Pemberton-Pigott crispinpigott at gmail.com
Thu Apr 26 17:00:10 CDT 2012


Dear David W

> Interesting comments Crispin.  Did no one listen to you in the Hague?

The Hague meeting was interesting and all the more interesting on that no
comment has been raised here about what took place there.

There were two general problems for everyone to handle. The first was that
there really was an expectation that the WBT 4.1.2 was going to be
rubber-stamped as the test to be used for placing stoves in a set of tiers.
As anyone who has used it knows, there is nearly no relationship between the
results of a set of such tests and performance in the field. This has been
said by all sorts of people: Dean Still, Penn Taylor, Nate Johnson, Rob
Baylis, you truly and so on. It is widely stated that in order to get any
idea what the fuel consumption and emissions will be one should conduct some
sort of field test.

So think about this: the test proposed, and assumed to become the
'international standard test' to place stoves on performance tiers was the
very test that pretty much all agree cannot do that particular task.

This was addressed in the following manner: two main thrusts were taken by
delegates. First, it was necessary to add wording that did not limit the
tests to be applied to the problem of differentiating between stoves. An
important 'removal' was the word and concept of 'simmering' which has no
scientific definition, is not defined in the document, and which is not a
standard metric in common use. Some stoves never simmer. That is just one
example. A second approach was to entrench the necessity to have a stove
testing protocol rated for precision in order to be able to state that the
rating of the stove is 'real'. There is no point in having a test method
that places a stove in three tiers simultaneously.

As people may or may not know, the regular Water Boiling Tests have not been
reviewed formally by an independent lab as suitable for the purpose for
which they are being employed. Penn Taylor reviewed the UCB WBT in at least
two or three papers and said it has systematic errors of about 50% in total.
To put a number on it, if you test repeatedly using the UCB WBT and average
the results, you can get a thermal efficiency number of 30%, plus or minus
15%. That is about the best you can get. That means it can place a stove on
a tier that is 45% from one to another. As you can imagine, that is not
going to fly very far.

As we do not know what the final text will look like (still with the
editorial committee) we have to wait. I submitted 18 suggestions for text
edits to remove some of the internal contradictions, in keeping with the
decisions of the group, but they 'don't work that way'. The text as edited
is the final text, basically. Of the 18, 13 were rejected, some accepted
partially and one or two accepted. It is not going to be internally
consistent. I do not know where that will leave the community as a whole.

One way forward is to find accurate ways to measure individual attributes of
the stoves. This can be done at fairly low cost but it will not be a 'water
boiling test' which contains so many steps each of which has large inherent
errors (systematic one).

It will be interesting of course to hear from others and what they think
will have to be done to place stoves on tiers.

Regards
Crispin






More information about the Stoves mailing list