[Stoves] Error in the recommended Shell Foundation HEH ProjectWater Boiling Testhttp://ehs.sph.berkeley.edu/hem/hem/protocols/WBT_data-calculation_sheet.xls

[Crispin Pemberton-Pigott]

Dear Dana

That is a sensible plan.  Andrew has used a very slightly different approach
to the calculation of the effect of moisture and arrived at essentially the
same result that Nigel and I got. I will reply to Andrew's post separately.

The spreadsheet is a very fine piece of work with its drop down lists and
all, it is an error on the moisture compensated energy figure that is the
problem.  In fact I am not saying we checked out the entire sheet though
that might be a good idea.  

If the sheet was used by Rob (congratulations Rob!) for his work presented
earlier this year, he might find this correction more significant than the
initial v.s. final water problem.  I didn't put actual numbers to it yet but
it cold mean that a stove that performed to a certain level in a lab with
dry wood might under-perform by 20 to 30% in the field using damp wood, even
when the moisture was taken into account (using the HEH spreadsheet for test
analysis).  In fact that under-performance might not be there.  It can
explain a lot of anomalous results.

Thanks for your attention to this.

>From Rob's article (p.60) " Variation in the data was reduced when
calculations were done using initial water volume, which is expected;
however, we found no change
in the relationship between TCS and ICS when both calculation methods are
compared."

It was the reduction in that variation I was looking for.  Thanks to Rob for
doing that recalculation. I have found that the use of the initial water
volume also reduces the possibility of getting a 'false positive' (better
than reality) comparison between two very similar stoves.  If you started
two simmers with two identical amounts of water, the only variable is the
wood.  If you take the same starting position and calculate with two
different final masses of both wood and water, misleading things creep in.

If the operators misbehaves (runs the water temperature too high) it reports
a higher consumption.  The operator can do a better job next time.  However
it is difficult to get a performance that is misleadingly high at any time.
It really shows improvements when tweaking a stove with a small change.

"Nor did we find any change in the (poor) relationship between lab-based
stove performance (SP) indicators and fuel consumption in the field."

It is that "poor relationship" to which I refer now.  Rob might recalculate
a set of data with the corrected cell and then let us know if the field
tests suddenly jump into line with the lab tests.

Thanks everyone
Crispin 

-----Original Message-----
From: Dana Charron [mailto:danacharron at yahoo.com] On Behalf Of
dana at ceihd.com
Sent: September 18, 2007 1:31 PM

Dear Crispin and Colleagues,

I can comment on the specific consumption issue as I discussed this with Rob
recently.  He noted that he had addressed this issue in the recent article
that appeared in "Energy for Sustainable Development."  He referred me to
page 4, left column, second full paragraph.  He offered to share his data
with anyone who is interested, and he noted that in his case at least, the
change in the specific consumption calculation did not add any explanatory
power to the data.

Perhaps Rob will weigh in directly, but as you may know, he has a brand-new
baby at home and so is rather swamped.  (Congratulations, Rob!)

The CEIHD website is in a transition right now, but as soon as we have the
new site functioning, we can make changes to the spreadsheet.  If we are
still lacking consensus on what the calculations should be, we will at a
mimimum post your concerns as an appendix to the document. 

--Dana

Dana Charron
Director, Household Energy and Health
Center for Entrepreneurship in International Health and Development (CEIHD)
School of Public Health, University California Berkeley
Tel: 510-547-4036 (Direct)
Tel: 510-643-6432
Fax: 510-643-8236
Email: dana at ceihd.com
http://ceihd.berkeley.edu/