[Stoves] [Ethos] Additional presentations at ETHOS 2015

[Crispin Pemberton-Pigott]

Dear Sam

Thanks for posting the presentation "How to Cheat on the WBT" addressing the
issue of low power performance measurement.

The core of the problem is not that changing the mass of water changes the
result of the calculation, giving a different rating for the performance on
that metric. That was always true. 

The root problem is that metric 'Specific Fuel Consumption' for low power is
not valid in the first place. Please see 

KEY DIFFERENCES OF PERFORMANCE TEST PROTOCOLS FOR HOUSEHOLD BIOMASS
COOKSTOVES, Yixiang Zhang, Crispin Pemberton-Pigott, Zongxi Zhang, Hongyan
Ding, Yuguang Zhou, Renjie Dong

DOI:10.1109/DUE.2014.6827753

It is one of the issues discussed.

It is remarkable to me that this has not been comprehensively addressed by
the US teams in the current environment of trying to improve the usefulness
of stove test results.  During the development of WBT 4.x the matter was
stonewalled, to everyone's disadvantage.

Before we get into the details of the mathematics or the gaming of a stove
test, the metric has to have validity in the first place. It doesn't.

Why?  Because there is no relationship between the heat needed to keep a hot
pot hot and the mass of water contained in it. Was explored at Eindhoven in
the early 80's. The SFC was rejected also by Rani et al in 1991, again in
2014 by Zhang et al. Zhang's experiment was reproduced by Jim Jetter last
year re-confirming that the mass of water has no influence in the amount of
fuel needed. What value can there be in dividing one number by the other?
Why not divide by 8 instead?

Accepting these multiple experiments, it is conceptually obvious that
changing the mass of water used to divide into the mass of fuel consumed (or
the energy number) renders the result meaningless.  Not 'less useful',
'meaningless'.  Zhang's experiment just did the experiment very accurately
(four 9's). The Indians did it quickly and pointed out that it supported the
conceptual analysis that low power specific fuel consumption was not a
useful metric because it didn't tell us anything.

Thank you for bring this to the attention of ETHOS. It has far-reaching
implications.

All three of the low power metrics in the IWA are have no physical basis and
therefore have no value for predicting anything. Part of the inherent 30%
variability in the WBT is caused by the use of poorly conceived metrics that
report 'numbers'. One of the motivations for the creation of Working Group 1
under ISO TC-285 was to assure the use of correctly grounded metrics 'with
value' for regulation and performance rating.

Debating how to influence the result of a SFC low power number is like
debating what colour to paint the hull of the Titanic.  The problem is that
of trying to assign an engineering performance number - energy per unit work
done - to a task that does not require the accomplishment of work - keeping
a pot hot.

Consider this. When the simmering phase starts, take the pot off and put it
into a well-made retained heat cooker in the next room.  Keep the fire going
in the stove. Measure the emissions. After 45 minutes, divide the mass of
fuel used for the fire (or the energy released or anything else you want) by
the number of litres in the pot in the other room which is still hot, of
course. Do the same with another stove.  Compare the results. What have you
learned?

The emissions from the stove are unaffected by the amount of water in the
pot sitting in the retained heat cooker. If it was still on the stove
instead of in the RHC, the number of litres would still not affect the pot's
'need for heat'. The only 'need' is to overcome losses. Those losses are not
measured. Remaining hot is the only prescribed task. 

If the RTC is extremely well insulated, the heat loss would be zero. That
means the efficiency of a perfect simmer (which is what a perfect RCT
represents) is, according, according to the WBT, 0%.  The better the stove
and the more controllable and the fire, the lower the calculated final SFC
number. This shows the silliness of the metric.  A better result is supposed
to result in a better rating not a worse one. Where that number is on 'tier
4' is meaningless.

'Simmering' is a task. 'Specific Fuel Consumption' is an energy efficiency
metric. They are not compatible. Calling it Low Power (as happened at the
IWA after the obvious was rejected) does not make it relevant to performance
rating. I hope you can convince some of your peers. This matter is
distracting us from far more important topics.

Best regards
Crispin


-----Original Message-----
From: ethos-bounces at vrac.iastate.edu [mailto:ethos-bounces at vrac.iastate.edu]
On Behalf Of Samuel Bentson
Sent: Saturday, February 14, 2015 06:49
To: Anderson
Cc: ethos at vrac.iastate.edu; marchese at colostate.edu; Discussion of biomass
cooking stoves
Subject: Re: [Ethos] [Stoves] Additional presentations at ETHOS 2015

Hi Paul,

I chose that title to garner interest. It worked! The point is to show how
to better standardize the WBT, and in case that is not possible, to show how
to be aware of how the results can be manipulated or misunderstood. Here are
the slides I threw together in a few minutes after deciding not to do my
original talk. It's mainly just some headings. We're working on a paper that
goes into detail about the water quantity used, it also talks about the
temperature of the water during simmer, and the firepower used to bring the
water to boil.  You're right that the quantity of water influences the
emissions metrics. The current WBT does have some language about not
comparing tests that were conducted with different amounts of water, but it
is vague. I wasn't aware of it when I gave the talk and I don't think other
testers I have spoken to are aware of it.

Sam