[Stoves] Selecting stoves based on data... was Re: kgharris stove test

[Crispin Pemberton-Pigott]

Dear Paul

 

Great message both in intent and content.

 

The difficulty of getting numerical performance for a stove that is relevant
to the stove selection process (really, it is a pre-selection process) is
underestimated. This is in part because the tech-heads have dominated the
stove development process, especially chasing heat transfer efficiency and
fuel savings at the expense of all other functions.

 

I almost said 'most other functions' but I can live with 'all' for
discussion purposes.

 

Too much has been sacrificed for 'good' performance, good in the opinion the
tech-heads.

 

So the soc-heads (social science bods) have been protesting for years. What
is important about the Indonesian CSI Pilot was that the engineering and
social science teams created new methods of evaluation in the lab that are
far better predictors of acceptance and actual performance without reducing
everything to thumb-suck opinions about 'what people want'. 

 

The cooking power, degree of control and cooking power per pot size dominate
the acceptance of a stove. All these can be reduced to numbers obtainable in
a lab setting.  Additional metrics are ignition time to high power, working
life, cost, and long term cost. Numbers are assignable.

 

Indonesian rural women are much more concerned about safety than anywhere I
have been in Africa. In fact their expectations are far more constrained
about surface temperature than anywhere else I have been. This can be
reduced to a number.

 

So the proper blending of the anthropological survey information (not a
sociological mass survey) has provided a really good impetus to this new
approach where the benefits of a stove undergoing a pre-selection process
can be realistically assessed before hitting the (much more expensive) focus
group practice sessions.

 

For the numbers, not many require additional measurements. We have had most
of the data all along. What we did with the CSI method was form the test
method around the required outputs, without those required outputs being
established by taking a fresh look at the market as seen through the eyes of
stove promoters, not designers.

 

Regards

Crispin

 

 

Crispin, Kirk and all,

Subject Line changed to reflect the change in the Subject.  This is NOT
about Kirk's stove, but about DATA about stoves being used for stove
selection.

Crispin's message (below) raises TOO MANY questions about the SPECIFICS OF
stove test data.  Instead, this message/topic is about what is actually done
when selecting stoves that would be put into the market in a project.   It
is not about the household or cook selecting a stove from the several
choices in the market or project.   It is about project leaders and
scientists examining stove characteristics when making selections.

My contention is that the numbers (data) do not matter all that much.
There are simply too many variables, so "perceptual averages" (or
generalities) come into the process.  One example is from (IMO = in my
opinion) one of the best selection efforts that use data, the large project
in Indonesia.  It created a 3-star qualitative system to examine (I believe)
3 major characteristics (like emissions, efficiency......).   Really great
implementation of a rating system linked to financial assistance for stove
purchases.  [[[ The much discussed Tier 0-to-4 system has not been used in a
similar way, as far as I know, even though there is a stove database that
gathers such data. ]]]

But even with those stars, in the end the factors of cost and attractiveness
and convience of use and fuel issues can simply overwhelm the "data"
approach.  

So I say (mainly to Crispin, but to all), the three numbers for fire-power
can be used as indicators, if used at all.   The "devil in the details"
(such as moisture content of the fuel, and including or excluding time and
fuel during ignition time, etc.) simply blow away the data themselves.  In
other words, even with all of the fullest calculations possible, words such
as "6 kW firepower" simply mean that this combustion arrangement "can really
put out some heat."   From that generality, stove designers work with pot
placement, skirts, plancha vs direct flame to pot, size changes, etc etc
etc.

For some readers (and project leaders, etc.), this message is terrible
because it diminishes the importance of data (and some related employment
and budgets).   Others see it as "that's what I do."   Others say "so what?
Not interested.  Already have too much invested in my stove and factory, so
I ignore the data evaluations."   

For stove INNOVATORS like Kirk Harris who are making experimental units,
getting test results are extremely important to fine-tune the improvements
so that we do (and should) pay attention to new developments.  

But in the end, the data do not mean acceptance or rejection of a
manufactured line of stoves.  Marketing overwhelms stove science.  So do
cost and conveinece, etc.

Paul




Doc  /  Dr TLUD  /  Prof. Paul S. Anderson, PhD
Email:  psanders at ilstu.edu <mailto:psanders at ilstu.edu> 
Skype:   paultlud    Phone: +1-309-452-7072
Website:  www.drtlud.com <http://www.drtlud.com> 

On 6/9/2016 7:01 AM, Crispin Pemberton-Pigott wrote:

Dear Kirk and Paul

 

There were three numbers provided for power, the low the high and the
test-high.

 

It was mentioned that there was a scale under the stove (which is a very
good idea) and that readings were taken per 5 minutes (which is a pretty
long interval).

 

Was the high power a maximum during the test phase, or the average over the
phase? If it was the average then it should be presented like that because
the phase includes an ignition. If it is the maximum firepower, that would
not be the average including ignition.

 

If I say an electric stove has a 5 kW element, that is the maximum input
power.  The cooking power is the input power times the heat transfer
efficiency.  That is pretty straightforward.

 

So I wondered if the scale readings were used to determine the maximum power
by using the burn rate, or if the power was determined in some other way. If
it was scale-based, it means it is the average of the 5 minute intervals. If
it was based on the scale, what energy content was applied? Was it the WBT
sheet LHV (compensated for moisture? Was it the As Received, Ash-Free
heating value? Was that value char-remaining-compensated to correct for the
energy not released? 

 

The determination of the firepower of a stove is not nearly as easy as
getting the cooking power, which in any case is what people experience when
they use it.

 

At the moment I do not have a clear picture of what that 6 kW refers to, and
as mentioned already, we don't know the cooking power (delivered energy).
Planning to introduce a design to a certain area will require taking a
pretty good guess that the product is in the right cooking power range.

 

Because we don't' have the test spreadsheet with all the input numbers it is
hard to know these answers, hence all the questions.

 

Dr Nurhuda provided a photo of a pot as proof that the PM is low for his
latest stove. That is not a 'test result' in the numerical sense, but we can
assume it must be pretty good.

 

Kirk provided some numbers from a WBT conducted at Aprovecho, which have
emissions per MJ and per unit time etc. But we don't have the numbers for
the total MJ or the total time so we can't tell a whole bunch of things we
need to know to evaluate the performance or suitability. 

 

The best we could do is to share the official test spreadsheet so the
missing metrics are available to those who need them. I cannot select a
stove design based on an emissions per MJ if I don't know how many MJ it is
capable of delivering. I can rate a stove I have already selected using that
metric.

 

I can't select a stove design based on a heat transfer efficiency if I don't
know how much heat was involved. Similarly, the firepower combined with
other metrics could be used to determine cooking capabilities but only if I
am quite sure what that firepower number refers to: peak, average, smoothed
3-minute peak, and so on. 

 

What do we want to know?
Cooking power total.

Cooking power per sq cm.

System efficiency for some known and relevant task, expressed as the "Energy
Efficiency"

Fuel flexibility.

Cost.

Durability.

Pot holding capacity.

Pot stability (when stirring).

Tilt stability.

PM mass per delivered MJ (used together with the cooking power)

CO mass per delivered MJ (ditto)

 

Note: the Energy efficiency is used as a proxy for relative fuel consumption
to compare stoves. Very importantly, it must not refer to the heat transfer
efficiency or any proxy of it. That is an internal relationship, and we want
the black box version of the fuel consumption.

 

Most of these numbers can be calculated from the information recorded in a
WBT spreadsheet.

 

Regards

Crispin

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20160609/d0c4f066/attachment.html>